Analysis of Shannon's entropy to contrast between the Embodied and Neurocentrist hypothesis of conscious experience.
Sergio J Martínez García, Pablo Padilla Longoria
Bio Systems December 1, 2024 DOI: 10.1016/j.biosystems.2024.105323 via PubMed
Summary
A proposed index based on Shannon's entropy can discriminate between the Neurocentrist hypothesis (consciousness is in the brain) and the Embodied hypothesis (consciousness involves the whole organism). Current theories like information integration theory and global workspace theory focus only on the brain, failing to account for embodied processing. The index measures whether information processing is primarily internal or external, and simulations with networks of varying internal and outer layers validated the index as unbiased. This offers a way to test which hypothesis better explains consciousness without relying on physical network structure.
Study at a glance
| Characteristics | Theoretical or philosophical paper Peer reviewed |
|---|---|
| Keywords | Embodied hypothesis Kolmogorov complexity Neural network Neurocentrist hypothesis Shannon's entropy |
| Citations | 7 |
| Key finding | An entropy-based index can discriminate between Neurocentrist and Embodied hypotheses of consciousness, validated through simulations with networks of varying internal and outer layers. |
Abstract
We usually accept that consciousness is in the brain. This statement corresponds to a Neurocentrist view. However, with all the physical and physiological data currently available, a convincing explanation of how consciousness emerges has not been given this topic is aborded by Anil Seth (2021). Because of this, a natural question arises: Is consciousness really in the brain or not? If the answer is no, this corresponds to the Embodied perspective. We cannot discriminate between these two points of view because we cannot identify how the organism processes the information. If we try to measure information processing in the brain, then the Neurocentrist view is unavoidable. For example, the information integration theory of Tononi's research group and the global work area theory developed by Dehaene and Baars, focus solely on the brain without considering aspects of Embodied vision (See Tononi, 2021; Dehaene, 2021). In this article, we propose an index based on Shannon's entropy, capable of identifying the leading processing elements acting: Are they mainly inner or external? In order to validate it, we performed simulations with networks accounting for different amounts of internal and outer layers. Since Shannon's entropy is an abstract measure of the information content, this index is not dependent on the physical network nor the proportion of different layers. Therefore, we validate the index as free of bias. This index is a way to discriminate between Embodied from Neurocentrist hypotheses.